Why 3D-printing is heavily underappreciated
--
In 2005, a group at Bath University started work on the RepRap project. It was an attempt at implementing a very interesting thought: Can we create a machine that can copy itself?
It turns out that you can (albeit only to a certain degree at this point in time). But much more importantly: the RepRap project showed that decentralized, open source development works not only for software, but also for hardware.
The RepRap project started out with a single design for a 3D printer made by Dr. Adrian Bowyer. It quickly grew as others became interested and started branching out from his design, creating their own versions and even experimenting with entirely different configurations. The result can be seen on the RepRap wiki, which hosts a very wide selections of all types of different printer designs that anyone can build and modify freely.
The nature of 3D-printers encourage decentralized development of hardware since they enable easy testing of new designs, but this also extends to any design that uses 3D-printed components. Therefore, it seems that there’s an unrealized potential in expanding the use of these manufacturing technologies.
On top of enabling fast design iteration cycles, using 3D-printed parts for product manufacturing is actually quite interesting. First of all, it’s very flexible compared to injection molding, albeit not as fast: you can digitally decide exactly which components you want to print, how many and in which quality as well as with what material.
Despite the fact that it’s slower per part, there are a lot of advantages: first of all, it’s quite simple to scale the production by simply adding more printers, which again, provides tremendous flexibility and allows vertical integration of the manufacturing process as well as enables a company to grow their capacity over time instead of having to make huge investments up front.
Second, since the printers are so cheap compared to the quality they can produce, you can actually reasonably compete with molded parts up to a decent volume. Particularly in cases where you want wide variety at low volumes, 3D-printing is approaching a state where it’s definitely feasible and cost-efficient if you keep it in mind during the design of your product.
Third, the fact that there’s no mold or other tooling means you can instantly iterate even on the designs you are manufacturing. This can be a huge advantage — one of the primary reasons hardware development and manufacturing is so scary is that you usually couldn’t just do a re-take. Once you ordered a hundred thousand units of something, you were committed. If the design had mistakes, you were screwed. This is no longer the case.
Fourth, the possible part complexity is increased, at least relative to the cost. You can do some very fancy things with injection molding, but it’s very expensive and requires a lot of consideration. With 3D-printing, you can often draw inspiration from the wide variety of open source designs and easily produce features such as nut traps, which can greatly simplify a design. The possibility of increased complexity can also help reduce part count in an assembly, since you can design one complex part to replace several other simpler parts.
Fifth, since 3D-printers have become available at a very low price point, virtually everyone has access to them. This is a perfect synergy for enabling micro-manufacturing and open source development. The practicality of decentralized development is drastically increased if anyone can simply print the parts, order a few components and build their own prototypes.
There’s a common sentiment that 3D-printed parts, at least those made on consumer-grade machines, simply aren’t really useful as mechanical parts for any serious application. I’m personally not exactly sure what this opinion is based on. Studies have been done on the characteristics of 3D-printed materials, and while you obviously can’t expect a part made on a consumer printer in ABS to rival a machined part in steel, the material properties really aren’t that bad. Neither is the surface quality, or the tolerances for that matter.
Tolerances are probably one of the main concerns, and rightfully so. 3D printing, at least at the consumer level, is not as repeatable as many other processes. Experience and processes to deal with this can help a lot. Solutions to this problem include using fairly extensive post-processing procedures to ensure that parts live up to spec. Depending on the requirements of a particular part, often times it might not even be critical.
Either way, the fact that companies like Prusa Research exist, and are running hundreds of their own printers using exactly the same 3D-printed parts successfully, is simply proof that we can’t draw conclusions without carefully examining the possibilities and experimenting with actually implementing additive manufacturing in real world scenarios.
I’m not saying 3D-printing is for everyone, or for everything. Still, despite the huge hype its received, particularly a few years ago, I believe it’s currently under-appreciated, and I encourage more people and companies to consider if they could benefit from this technology that is already so easily available.
If what you need is a super detailed part or mirror surface finish, then those particular parts are probably not the easiest to start with. The same is true if you need really large parts.
But if you’re building smaller, functional things, particularly in robotics and similar areas, and even more so if you’re doing open source projects, experimenting with 3D-printing for more than just mock-ups could prove to be very beneficial.
Earlier, I mentioned Prusa Research. Founded by one of those people who started contributing to the RepRap project early on, Josef Prusa. It has grown to a huge manufacturing and development company, currently shipping thousands of consumer-grade 3D printers a month based. These are based on a particular RepRap design that Prusa himself has laid name to for about a decade at this point. They are somewhat unique in the fact that they use 3D-printed components for a very large part of their products. Many companies emerged from the RepRap project, including Makerbot, Ultimaker, Lulzbot and many others, but Prusa Research seems special.
Prusa Research is probably one of my favourite companies. Their commitment to open source and their willingness to experiment with doing things in a new way is really inspiring, and the fact that they’re currently valued at almost $250 million goes to show that there’s not just value for the people who get to buy these great printers, but also a lot of profit to be made.
If you haven’t already seen it, search “Prusa printer farm” on YouTube to see their huge array of their own printers, rapidly replicating the parts required to build new ones for customers. That is freaking cool.
What’s even cooler is visiting their Github page (https://github.com/prusa3) and seeing that they make all their design files available, even for the newest version of their printer, including their innovative heat bed design, using a PEI-coated steel sheet, attached magnetically to allow easy print removal and optimal adhesion.
That a company can be profitable, with thousands of customers, a large manufacturing facility, a valuation in the hundreds of millions of dollars, and still publish everything you need to replicate their product openly on Github, speaks volumes about the potential for a new type of industry in which we don’t hide behind patents and proprietary bullshit.
Imagine if more companies would follow suit, and how that would change the world. I don’t see a clearer path to a better future.
I recently wrote an article that provides a more high-level discussion of the potential implications of open source in development and organisations. You can read it here:
Please comment, clap, or send me an email at magnusnm@gmail.com; I want to hear what you think!
